AQUATIC CONSERVATION: MARINE AND FRESHWATER ECOSYSTEMS Aquatic Conserv: Mar. Freshw. Ecosyst. (2009) Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/aqc.1088

Potential impacts of projected sea-level rise on sea turtle rookeries

MMPB FUENTESa,Ã, CJ LIMPUSb, M HAMANNc and J DAWSON aSchool of Earth and Environmental Sciences, James Cook University, bQueensland Environmental Protection Agency, Brisbane, Australia cSchool of Earth and Environmental Sciences, James Cook University, Australia

ABSTRACT 1. Projected sea-level rise (SLR) is expected to cause shoreline erosion, saline intrusion into the water table and inundation and flooding of beaches and coastal areas. Areas most vulnerable to these physical impacts include small, tropical low-lying islands, which are often key habitat for threatened and endemic species, such as sea turtles. 2. Successful conservation of threatened species relies upon the ability of managers to understand current threats and to quantify and mitigate future threats to these species. This study investigated how sea-level rise might affect key rookeries (nesting grounds) (n 5 8) for the northern (nGBR) green turtle population, the largest green turtle population in the world. 3. 3-D elevation models were developed and applied to three SLR scenarios projected by the IPCC 2007 and an additional scenario that incorporates ice melting. Results indicate that up to 38% of available nesting area across all the rookeries may be inundated as a result of SLR. 4. Flooding, as a result of higher wave run-up during storms, will increase egg mortality at these rookeries affecting the overall reproductive success of the nGBR green turtle population. Information provided will aid managers to prioritize conservation efforts and to use realistic measures to mitigate potential SLR threats to the nGBR green turtle population. Copyright r 2009 John Wiley & Sons, Ltd.

Received 13 January 2009; Revised 13 June 2009; Accepted 8 October 2009

KEY WORDS: sea turtles; green turtles; sea-level rise; habitat loss; climate change; reef island; Great Barrier Reef;

INTRODUCTION 1993; Fish et al., 2005, 2008; Baker et al., 2006; LaFever et al., 2007). This is expected to cause a plethora of biogeophysical Sea level is anticipated to rise significantly in the future, with a and socio-economic consequences producing a cascade of projected sea-level rise (SLR) of 18 to 59 cm by 2100, and a impacts (Klein and Nicholls, 1999). Assessments of the possible additional 10 to 20 cm increase from melting ice sheets impacts of projected SLR at areas of high human population and glaciers (Overpeck et al., 2006; McInnes and O’Farrell, density, economic importance and/or areas that have high 2007; Meehl et al., 2007). Small, tropical low-lying islands, environmental value (e.g. areas important for threatened especially those that are not vegetated or lie on exposed reefs in species), can aid resource management planning and areas of high tidal range, are the most vulnerable to SLR conservation of wildlife that rely on areas at risk (Baker (Woodroffe et al., 1999; Church and White, 2006). Impacts et al., 2006; Cowell et al., 2006). anticipated from SLR include saline intrusion into the water Currently, concerns exist regarding the impacts of SLR on table as well as inundation and flooding of beaches and the most important rookery, Raine Island, and several of the shoreline erosion of coastal areas (Klein and Nicholls, 1999; smaller cays (e.g. Bramble ) used by the largest green turtle Mimura, 1999). Previous studies indicate that the most (Chelonia mydas) population in the world: the northern Great significant impacts will be at residential and recreational Barrier Reef (nGBR) green turtle population. This population areas, agricultural land (Nicholls, 2002; Snoussi et al., 2008), nests at rookeries in the far nGBR and Torres Strait region, wetlands (Nicholls et al., 1999; Nicholls, 2004) and habitats for with an average of 50 000 turtles on a high nesting year threatened, endangered and endemic species (Daniels et al., (Limpus et al., 2003). Over the last 10 years a reduction in

*Correspondence to: MMPB Fuentes, School of Earth and Environmental Sciences, James Cook University, Australia. E-mail: [email protected]

Copyright r 2009 John Wiley & Sons, Ltd. MMPB FUENTES ET AL. hatching success has been observed at Raine Island, which is thought to be caused by rising groundwater and other geomorphic processes (e.g. movement of sand) (Limpus et al., 2003). It is believed that SLR is likely to exacerbate these processes and the frequency of nest inundation (Limpus et al., 2003). The present study uses a geographic information system (GIS) to map the impacts of projected SLR, in terms of inundated area, under four SLR scenarios on a selection of sites used for nesting by this threatened population. The impacts of SLR on sea turtle nesting grounds has previously been quantified in Bonaire and Barbados (Fish et al., 2005, 2008), the east coast of the USA (Daniels et al., 1993) and the Hawaiian Islands (Baker et al., 2006). However, there has been no study in Australia, an area that contains globally significant marine turtle populations. In addition, prior studies, with the exception of Baker et al. (2006), focus on the impacts to only one rookery for a particular turtle population. Such an approach does not provide a full understanding of how a genetic stock (management unit) will be affected and respond to SLR. Since sea turtles may shift nesting grounds when nesting habitat is no longer available (Hamann et al., 2007) there is also the need to investigate how a variety of nesting grounds for the same population will be affected. Considering this, the impacts of SLR on eight different sites in north-east Australia, representing 99% of nesting activity for the nGBR green turtle population (Limpus et al., 2003), were investigated. This ensures that managers will be able to direct and focus management and conservation actions strategically to protect the nGBR green turtle population Figure 1. Location of rookeries utilized by the nGBR green turtle from impacts of SLR. The ecological impacts of loss and population and their relative importance as a nesting ground. alteration of nesting habitat on the nGBR green turtle population are also discussed. Characteristics of rookeries Beach profiles were measured at Bramble Cay, Dowar Island (north, south and west beaches) and Milman Island relative to METHODS low water mark at 100 m intervals (except at Dowar where a 50 m interval was conducted), using the dumpy level standard surveying technique (Mwakumanya and Bdo, 2007), where Rookeries elevation of points (z) along the transect are calculated from Nesting by the nGBR green turtle population occurs, in order slope and ground distances. Waypoints (x and y) were of importance, at : (1) Raine Island (111360S, 1441010E), (2) recorded at each elevation point from the profile transects Moulter Cay (111260S, 1441000E), (3) Bramble Cay (91090S, using a global positioning system (GPS) as well as bearings. 1421530E),(4) Dowar Island (91550S, 1441020E), (5) MacLennan The x, y and z coordinates for each point from the beach Cay (111220S 1431 480E), (6) Sandbank 7 (131260S; 1431580E), profiles were used to construct triangulated irregular network (7) Sandbank 8 (131210S; 1431570E) and (8) Milman Island (TIN) models for each beach using the 3-D analyst tool in s (111100S; 1431 000 E) (Figure 1). The highest concentration of ArcGIS . Data from Raine Island were collected using Real nesting occurs at Raine Island and Moulter Cay, with 90% of Time Kinematic (RTK) GPS. Beach width, mean and the overall nesting for this population (Limpus et al., 2003). maximum elevation values and area available for nesting for Smaller, but significant, numbers of nests are laid at both each beach were obtained from the TIN models. Beach profiles Bramble Cay and Dowar Island in Torres Strait. Nesting at for Moulter Cay, MacLennan Cay, Sandbank 8 and 7 were Dowar occurs at three distinct beaches: the north, south and derived from existing information on their elevation profiles west beaches. Lowest nesting density occurs at Milman Island and morphology (King and Limpus, 1983; King et al., 1983a, (Dobbs et al., 1999). Raine Island, Moulter Cay, MacLennan 1983b). Spatial information for Moulter Cay was obtained Cay, Sandbank 7 and Sandbank 8 are non-vegetated or from an aerial photograph taken in 1990 (0.25 m pixel vegetated sand cays and Milman Island is a forested cay, resolution). located in the nGBR (Figure 1). Bramble Cay is located on a fringing reef around a small volcanic outcrop in the north-east Nesting activity of Torres Strait. Dowar Island is one of the Mer group islands, which are volcanic high islands fringed with Holocene reef. All Surveys of nest location were carried out to determine the spatial rookeries are small in size (0.02 to 0.3 km2). distribution of nests and the preferred nesting habitat–in terms of

Copyright r 2009 John Wiley & Sons, Ltd. Aquatic Conserv: Mar. Freshw. Ecosyst. (2009) DOI: 10.1002/aqc POTENTIAL IMPACTS OF PROJECTED SEA-LEVEL RISE ON SEA TURTLE ROOKERIES elevation and distance from high water mark (HWM)–at each from an intermediate sea-level rise scenario (0.35 m) was rookery. Owing to logistical and time constrains, surveys for turtle considered as a measure of vulnerability (modified method nests were carried out only at Bramble Cay, Dowar Island, from Fish et al., 2005), and Pearson’s Correlation was used to Milman Island and Raine Island. Monitoring occurred during the examine the effects of each physical attribute and vulnerability 2006/2007 nesting season, which was a high nesting season with up to SLR. to 21 000 turtles nesting per night at Raine Island (CJL, unpublished data). Monitoring at Raine Island was conducted Threat to nesting area during storm events by Parks and Wildlife (QPW) as part of their annual monitoring programme, which has taken place since the 1970s As it is anticipated that waves will penetrate even further (Limpus et al., 2003). Turtles nested on all available un-vegetated inland during episodic storms (Gornitz, 1991; Fletcher III, beach area and therefore this study assumes that turtles nested 1992; Church et al., 2006), it was also explored how nests and everywhere above HWM and below the cliff line and outside any nesting areas will be impacted during storms under an central rock area. Nesting activities at Dowar and Milman Islands intermediate SLR scenario of 0.35 m rise. Due to lack of were monitored for 10 days during peak nesting and nest locations storm tide predictions previous highest astronomical tide were recorded with a GPS (Garmin Etrex, Garmin International, (HAT) measurements were used as an indication of possible Inc. Kansas). Nesting at Bramble Cay was monitored for a single intrusion by storm-wave run-up. Using data from the day during the 2006/2007 nesting season, therefore nesting Environmental Protection Agency, Australian Bureau of information collected during the 2007/2008 season was also used Meteorology website (http://www.bom.gov.au/index.shtml) as an indication of the location of nests at this site. The preferred and Seafarer tides, HAT was calculated to be 1.0 and 0.45 m elevation range, where 470% of nesting takes place, was above mean spring high tide level in Torres Strait and the calculated for each of the rookeries for which nesting nGBR region correspondingly. As HAT data are only information was available, by using zonal statistics (ArcGIS 9.0). available at a regional level, these are used only as an indicative measurement. It was then assumed that nesting Sea level scenarios and threat to nesting area area under 1.35 m (0.35 m SLR11.0 m run-up) and 0.8 m (0.35 m SLR10.45 m run-up) in Torres Strait and nGBR, Three SLR scenarios (0.18, 0.35 and 0.59 m) were considered respectively, would be affected by wave run-up during storm for 2100, from the IPCC 2007 (Meehl et al., 2007) and an events and consequently the nests laid in this area would be additional scenario (0.79 m) that accounted for ice melting into inundated. the system (0.2 m added to the highest scenario from the IPCC 2007 (Overpeck et al., 2006; McInnes and O’Farrell, 2007). Similar to other studies (Fish et al., 2005) impacts through RESULTS inundation of the nesting area were considered. For this, the TIN models were used to identify nesting area below each of Rookeries characteristics and nesting activity the elevations (0.18, 0.35, 0.59 and 0.79 m) and therefore areas Raine Island, Moulter Cay, Milman Island and north Dowar that would be inundated by SLR. The area inundated was provide the largest available nesting areas, and conversely, measured from the HWM. Analyses were conducted using the western Dowar provides the smallest area for turtle nesting Surface Volume tool in the ArcGis 9.0 - 3D Analyst Toolbox. (Table 1). The highest elevations were found at north Dowar and Raine Island (9.13 m and 4.9 m, respectively), while the Predicting threat to rookeries where beach profiles were lowest nesting beaches were at Sandbank 7, MacLennan Cay, not conducted Sandbank 8 and west Dowar (Table 1). Preferred nesting Owing to logistical constraints it was not possible to measure habitat varied at each rookery (Table 2), turtles at north beach profiles at Moulter Cay, MacLennan Cay, Sandbanks 8 Dowar nest at higher elevation and turtles at west Dowar nest and 7. To calculate the probable inundation at these rookeries, at lower elevations (Table 2). Preferred nesting elevation was it was first examined if there was a significant correlation found to be a result of the elevation range found at each between the maximum elevation at each rookery where a rookery, as the mean nest elevation was significantly and beach profile was conducted and the percentage of area lost for positively correlated with maximum and mean elevation at every SLR scenario. After this relationship was established a linear regression model was created to predict the probable Table 1. Characteristic of rookeries during the 2006/2007 nesting percentage of area inundated for the rookeries where profiles season. Rookeries are listed in order of importance were not conducted. To validate the predictive efficiency of the Rookery Width Nesting Mean Maximum linear model created, paired - t tests were run with the values of (m) area (m2) elevation (m) elevation (m) percentage of lost area calculated from the beach profile models with the values generated from the linear model for the Raine Island 90 152 247 1.2 4.9 field study sites (Raine Island, Bramble Cay, Dowar Island Moulter Cay N/A 78 200 N/A 3.0 Bramble Cay 44.4 21 980 1.34 4.08 and Milman Island). North Dowar 37.5 36 719 2.36 9.13 South Dowar 28 8803 1.021 3.9 Vulnerability as a result of rookeries characteristics West Dowar 19.4 3844 0.77 2.09 MacLennan Cay 38 24 000 N/A 1.05 The relationship between threat to nesting area and different Sandbank 7 45 22 000 N/A 0.76 physical attributes of each rookery (i.e. beach width, nesting Sandbank 8 60 32 000 N/A 1.36 area as well as maximum and mean elevation) was also Milman Island 17 58 648 1.8 4.28 investigated. For this, the proportion of beach under threat N/A 5 not available.

Copyright r 2009 John Wiley & Sons, Ltd. Aquatic Conserv: Mar. Freshw. Ecosyst. (2009) DOI: 10.1002/aqc MMPB FUENTES ET AL. each rookery (r2 5 0.959, n 5 5, P 5 0.01 and r2 5 0.989, n 5 5, Vulnerability and rookeries characteristics P 5 0.001, respectively). Mean distance of nest to HWM also Between 8% and 38% of the total area available for nesting varied between rookeries (Table 2), with mean nest distance (438 441 m2) across the beaches studied are predicted to be being positively correlated with beach width (r2 5 0.855, n 5 5, inundated under the various SLR scenarios (Figure 2). P 5 0.001). Sandbank 7 is predicted to lose the greatest amount of beach (12 to 49%), followed by west Dowar, MacLennan Cay and Threat to nesting area Sandbank 8 where approximately 11 to 47% of their area is predicted to be lost. Similarly, Milman Island is predicted to Validation of methods for rookeries where beach profiles lose 10 to 42% of its nesting area. North Dowar is predicted to were not conducted be the least vulnerable rookery with a predicted area inundated of 3–15% (Figure 2). As a significant correlation existed between the maximum Beaches with lower elevation were, not surprisingly, found elevation at each rookery and percentage of area lost for every to be more susceptible to SLR as inundation was significantly SLR scenario, for the beaches for which there were beach and negatively correlated with maximum elevation under all 2 profiles (except scenario 3) (scenario 1, r 5 0.831, n 5 6, SLR scenarios (scenario 1, r2 5 0.91, n 5 10, Po0.000; 2 P 5 0.041; scenario 2, r 5 0.842, n 5 6, P 5 0.035; and scenario 2, r2 5 0.91, n 5 10, Po0.000; scenario 3, 2 scenario 4, r 5 0.863, n 5 6, P 5 0.027), a linear regression r2 5 0.88, n 5 10, P 5 0.001 and scenario 4, r2 5 0.92, model was created to predict the probable percentage of area n 5 10, Po0.000) (Figure 3). inundated for the beaches for which it was not possible to measure profiles (Moulter Cay, MacLennan Cay, Sandbank 8 Threat during storm events and 7). Paired t-tests validated the linear models, as there was no significant difference between the values from the beach During storm events the nesting habitat at west Dowar is profiles and the values calculated from the linear models for all predicted to be under the greatest threat, with up to 75% of four SLR scenarios (all pairs, t 5 0.001, df 5 5, P40.999). available nesting habitat inundated, potentially affecting 90%

Table 2. Characteristic of preferred nesting habitat at each rookery during the 2006/2007 nesting season Rookery Preferred elevation Mean nest Percentage Mean distance range (m) elevation (m) of nesting at from HWM (m) preferred nest elevation

Bramble Cay 1.5–3.5 2.1 77.0 19.6 North Dowar 2.5–4.5 3.3 73.4 24 South Dowar 1.0–2.5 1.6 82.1 13 West Dowar 1.0–2.0 1.2 71.4 8 Milman Island 2.0–3.5 2.4 75.2 11.5 Rookeries are listed in order of importance. Data for Raine Island, Moulter Cay, MacLennan Cay, Sandbank 7 and 8 are not available. Preferred elevation range is where 470% of nesting takes place at each rookery. Elevation is measured from high water mark (HWM).

Figure 2. Area predicted to be inundated after sea-level rise scenarios of (1) 0.18 m rise, (2) 0.35 m rise, (3) 0.59 m rise, and (4) 0.79 m rise at each rookery used by the nGBR green turtle population.

Copyright r 2009 John Wiley & Sons, Ltd. Aquatic Conserv: Mar. Freshw. Ecosyst. (2009) DOI: 10.1002/aqc POTENTIAL IMPACTS OF PROJECTED SEA-LEVEL RISE ON SEA TURTLE ROOKERIES of nests laid. Milman Island, Moulter Cay, Bramble Cay, scenarios. It is predicted that under the most extreme SLR MacLennan Cay and Sandbank 8, Sandbank 7 and Milman scenario proposed by IPCC (2007)–a 0.59 m rise–27.7% of the Island are also predicted to have large amounts (450%) of total nesting area available for the nGBR green turtle their nesting area inundated during storm events, with population could be inundated. The extent of inundation of Bramble Cay and Milman Island potentially having up to individual beaches ranges from 11% to 36%, with the beaches 30% of their nests inundated. Raine Island is expected to have that support the highest levels of nesting being the least up to 30% of the available nesting area inundated during vulnerable to inundation. Similar results are predicted for sea storm events (Figure 4). turtle rookeries in the Caribbean region where 26% and 32% of the nesting area in Barbados and Bonaire, respectively, are predicted to be inundated with a 0.5 m (Fish et al., 2005, 2008). DISCUSSION Reduction of available nesting area will amplify density- dependent issues at nesting grounds, potentially increasing nest Threat of sea-level rise infection (Fish et al., 2008) and destruction of nests by co- To successfully conserve and manage sea turtles as climate specifics (Bustard and Tognetti, 1969; Girondot et al., 2002; change progresses managers will need to identify, understand, Limpus et al., 2003). This already occurs at Raine Island, predict and mitigate any future impact on these endangered Moulter Cay, Dowar and Bramble Cay during high density species (Hamann et al., 2007). This study quantified the area of nesting years. Higher nesting density at a particular rookery rookeries, utilized by the largest green turtle population in the may also reduce the total reproductive output as increased world that will potentially be susceptible to projected SLR disturbance by nesting co-specifics could result in premature use of somatic energy stores and resorption of ovarian follicles (Hamann et al., 2002; Limpus et al., 2003). Another outcome of SLR is increased impact of storm events, causing periodic beach erosion and washing away and flooding of nests (Gornitz, 1991; Fletcher III, 1992; Church et al., 2006). This will increase egg mortality affecting the overall reproductive success of the nGBR green turtle population. Further flooding of sea turtle nests and impacts on the reproductive output of sea turtles can occur through a raised water table as a result of SLR (Titus et al., 1991; Ross et al., 1994). Raine Island, in particular, is more susceptible to this as it already experiences water level problems (Hamann et al., 2007). On occasion, groundwater level has been so high at Figure 3. Relationship between maximum elevation at each rookery Raine Island that pooled water has been observed in and area predicted to be inundated after sea-level rise scenarios of (1) depressions and body pits made by turtles (Limpus et al., 0.18 m rise, (2) 0.35 m rise, (3) 0.59 m rise, and (4) 0.79 m rise. 2006). A recent study by Guard et al. (2008) has initiated

Figure 4. Percentage of nesting area and nests inundated, at rookeries for the nGBR green turtle population, under an intermediate sea-level rise of 0.35 m both during storms (S) and when storm events are not occurring (NS). Information on nest inundated is only available for rookeries where we conducted nesting monitoring.

Copyright r 2009 John Wiley & Sons, Ltd. Aquatic Conserv: Mar. Freshw. Ecosyst. (2009) DOI: 10.1002/aqc MMPB FUENTES ET AL. exploration of the water table dynamics at Raine Island in 2004; Fish et al., 2005). The most common and widely used order to provide models of water table response to tidal model is the ‘Bruun rule’ (Bruun, 1962) (see Cooper and Pilkey oscillations. Further investigation and expansion of this study (2004) for a compiled list of studies), which assumes a may provide more quantitative insights into the impacts of continuous equilibrium of sand transport between beach and SLR on groundwater dynamics and therefore the impact of nearshore (Woodroffe, 2008) and therefore it is not applicable SLR on the reproductive output at Raine Island. for the systems studied here. In addition, this model has been criticized for its restrictive assumptions, omission of important Possible responses by turtles and consequences of SLR variables and erroneous concepts (Cooper and Pilkey, 2004). To overcome some of the issues Cowell et al. (2006) recently Sea turtles may be able to adapt to SLR by shifting nesting up suggested incorporating probabilistic components to model the beach, away from the high tide (Fish et al., 2005; Limpus, outputs to allow greater freedom in quantifying some of the 2006). However, such a shift is constrained at small low-lying input parameters; however, owing to lack of specific data, islands and where urban development restrains landward especially on the coastal processes and changes in beach beach recession (Fish et al., 2008). As the nGBR green turtle profiles at each nesting ground, this model could not be population nests on beaches where little urban development applied. exists, a landward shift in nesting is a potential response for As assessing the quantitative impacts of shoreline erosion, this population. This is not the situation, however, for rise of water table and potential accretionary events was populations nesting at beaches developed for tourism in the beyond the scope of this study and therefore not incorporated Caribbean region (Fish et al., 2005, 2008). into the results presented here, it is important to consider that Some nesting beaches may become fully inundated as sea influences from these factors could lead to greater or lesser level continues to rise above 1 m beyond 2100 (Turner and habitat loss. Several other studies (Daniels et al., 1993; Fish Batianoff, 2007). Turtles nesting at these or at beaches that et al., 2005; LaFever et al., 2007) have used a similar approach have no more elevated nesting habitat, as may occur at west to this study, and only quantified the impact of SLR caused by Dowar and Sandbank 7, will need to seek out new nesting sites inundation. As with many other predictions of beach response (Limpus, 2006; Hamann et al., 2007). For example, if nesting is to SLR, the current approaches include uncertainties (Cowell no longer possible at west Dowar, higher density nesting may et al., 2006). In this study uncertainties arise from (1) predicted occur at the southern and northern beaches–which provides SLR scenarios, (2) assumptions of how beaches will respond to more suitable habitat–or turtles may shift to nest at nearby SLR (in terms of their sea level and wave climate), and (3) the Mer or Waer Islands. There is also the possibility that turtles models used to quantify the impacts and response of SLR to will shift their nesting to new beaches that may develop/or selected beaches (Cowell et al., 2006). Possible errors from stabilize in the region as a result of SLR (Hamann et al., 2007). these uncertainties were minimized by (1) utilizing a range of Nest placement has been shown to affect hatchling success SLR scenarios consistent with IPCC 2007 as well as and sex ratio in the GBR green turtle population (Miller and incorporating possible increases in SLR through ice/glazier Limpus, 1981; Morreale et al., 1982) and any shift in their melting and increase in wave-run up, and (2) by using similar rookeries may influence this. Changes in nesting locations may assumptions and methodology to other studies that address also have severe implications and cause further conservation comparable questions. Nevertheless, the results presented here challenges if they are forced to nest where even fewer provide the first insights and the best current available conservation measures are in place or management is assessment of the potential effects of SLR on the nGBR logistically difficult. Conversely, changes may result in green turtle rookeries. Studies of this nature, which assess the improved population performance as turtles may start potential impacts of SLR on endangered megafauna, are nesting in areas with more favourable nesting and incubating extremely important, as they can potentially aid managers to condition and/or areas with less anthropogenic threats, such as prioritize management efforts and to use realistic measures to traditional hunting of turtle meat and eggs as occurs at the mitigate potential SLR threats to these ecologically important rookeries in Torres Strait. Longer-term consequences species. Some potential management measures to mitigate the associated with changes in nesting distribution include the impacts of inundation and erosion from SLR include (1) ‘hard development of new genetic stocks and thus differentiation in engineering structures’ (e.g. seawalls, groynes), (2) ‘soft biological parameters (e.g. turtle stocks with different breeding methods’ (e.g. beach nourishment, dune building), and (3) phenology and different size adult females) (Limpus, 2008). retreat and setback regulations (Nicholls and Tol, 2006; Fish This has been suggested to have occurred with the historical et al., 2008). In order to determine the most realistic and efficient (Pleistocene) population of flatbacks, Natator depressus, which solution to use, a cost benefit analysis of each strategy will be developed into two distinct current (Holocene) populations necessary as well as information on any ethical, ecological and (Limpus, 2008). practical issues associated with implementing them.

Further impacts of SLR and uncertainties ACKNOWLEDGEMENTS Turtle nesting beaches may be further affected by SLR through shoreline erosion, which is dependent on a series of This work was supported and funded by the Australian factors such as wave energy, tidal currents, island and reef Government’s Marine and Tropical Sciences Research morphology, sediment type and sediment supply, among Facility, Queensland Environmental Protection Agency others (Cooper and Pilkey, 2004; Woodroffe, 2008). (QEPA), including the Queensland Parks and Wildlife, and Developing appropriate models to successfully predict Torres Strait Regional Authority. We thank the traditional shoreline response to SLR is challenging (Cooper and Pilkey, owners of the field sites for allowing access and granting

Copyright r 2009 John Wiley & Sons, Ltd. Aquatic Conserv: Mar. Freshw. Ecosyst. (2009) DOI: 10.1002/aqc POTENTIAL IMPACTS OF PROJECTED SEA-LEVEL RISE ON SEA TURTLE ROOKERIES permission to work on their land. We also would like to thank Guard P, Macpherson K, Mohoupt J. 2008. A Field the staff and volunteers from QEPA, Earthwatch and Erub Investigation into the Groundwater Dynamics of Raine community for help conducting beach profiles. Special thanks Island. Division of Civil Engineering, the University of go to Moses Wailu, Ian Bell and Kenny Bedford for Queensland, Brisbane. facilitating work at Dowar Island, Milman Island and Hamann M, Limpus CJ, Whittier JM. 2002. Patterns of lipid storage and mobilisation in the female Bramble Cay, respectively. We are also grateful to Kirstin (Chelonia mydas). Journal of Comparative Physiology B Dobbs for helping to acquire elevation data on the rookeries Biochemical Systemic and Environmental Physiology 172: for which beach profiles were not available. Field work 485–493. complies with the current laws of Australia and all permits Hamann M, Limpus CJ, Read M. 2007. Vulnerability of necessary for the project were obtained (QPWS permit number marine reptiles in the Great Barrier Reef to Climate WISPO4316207). Change. In Climate Change and the Great Barrier Reef: A Vulnerability Assessment, Johnson J, Marshall P (eds). Great Barrier Reef Marine Park Authority and Australian Greenhouse Office: Townsville; 667–716. REFERENCES King BR, Limpus CJ. 1983. Sandbank No. 7, Great Barrier Reef, Queensland. Corella 7: 78–79. Baker JD, Littnan CL, Johnston DW. 2006. Potential effects King BR, Limpus CJ, Seton DHC. 1983a. Sandbank No. 8, of sea level rise on the terrestrial habitats of endangered and Great Barrier Reef, Queensland. Corella 7: 76–77. endemic megafauna in the Northwestern Hawaiian Islands. King BR, Limpus CJ, Seton DHC, Tomes GR. 1983b. Pandora Endangered Species Research 4: 1–10. Cay, Great Barrier Reef, Queensland. Corella 7: 71–73. Bruun P. 1962. Sea level rise as a cause of shore erosion. Klein RJT, Nicholls RJ. 1999. Assessment of coastal Journal of the Waterways and Harbors Division 88: 117–130. vulnerability to sea-level rise. Ambio 28: 182–187. Bustard HR, Tognetti KP. 1969. Green sea turtles: a discrete LaFever DH, Lopez RR, Feagin RA, Silvy NJ. 2007. simulation of density-dependent population regulation. Predicting the impacts of future sea-level rise on an Science 163: 939–941. endangered lagomorph. Environmental Management 40: Church J, Hunter JR, Kathleen LM, White NJ. 2006. Sea-level 430–437. DOI: 10.1007/s00267-006-0204-z. rise around the Australian coastline and the changing Limpus CJ. 2006. Impacts of climate change on sea turtles: a frequency of extreme sea-level events. Australian case study. In Roundtable on Migratory Species and Climate Meteorological Magazine 55: 253–260. DOI: 10.1016/ Change, CMS COP; Nairobi, November 2005. j.gloplachs.2006.04.001. Limpus CJ. 2008. Adapting to climate change: a case study of Church JA, White NJ. 2006. A 20th century acceleration in the flatback turtle, Natator depressus. Proceedings of the 27th global sea-level rise. Geophysical Research Letters 33. Annual Symposium on Sea Turtle Biology and Conservation. L01602. DOI: 10.1029/2005GL024826. http://www.nmfs.noaa.gov/pr/pdfs/species/turtlesymposium Cooper JAG, Pilkey OH. 2004. Sea-level rise and shoreline 2007.pdf retreat: time to abandon the Bruun Rule. Global and Limpus CJ, Miller JD, Parmenter CJ, Limpus DJ. 2003. The Planetary Change 43: 157–171. DOI: 10.1016/j.gloplacha. green turtle, Chelonia mydas, population of Raine Island and 2004.07.001. the northern Great Barrier Reef: 1843–2001. Memoirs of the Cowell PJ, Thom BG, Jones AR, Everts CH, Simanovic D. Queensland Museum 49: 349–440. 2006. Management of uncertainty in predicting climate- Limpus CJ, Limpus DJ, Munchow M, Barnes P. 2006. change impacts on beaches. Journal of Coastal Research 22: Queensland turtle conservation project: Raine Island turtle 232–245. study, 2004–2005. In Conservation Technical and Data Daniels RC, White TW, Chapman KK. 1993. Sea-level rise: Report, Jones M (ed.). Brisbane; 1–41. destruction of threatened and endangered species habitat in McInnes K, O’Farrell S. 2007. Regional climate change South Carolina. Environmental Management 17: 373–385. projection: sea level rise. In Climate Change in Australia, Dobbs KA, Miller JD, Limpus CJ, Landry AMJ. 1999. CSIRO: Australia; 92–97. Hawksbill turtle, Eretmochelys imbricata, nesting at Milman Meehl GA, Stocker TF, Collins WD, Friedlingstein P, Island, northern Great Barrier Reef, Australia. Chelonian Gaye AT, Gregory JM, Kitoh A, Knutti R, Murphy JM, Conservation and Biology 3: 344–361. Noda A, et al. 2007. Global climate projections. In Climate Fish MR, Cote IM, Gill JA, Jones AP, Renshoff S, Watkinson Change 2007: The Physical Science Basis. Solomon SD, AR. 2005. Predicting the impact of sea-level rise on Qin M, Manning Z, Chen M, Marquis KB, Averyt M, Caribbean sea turtle nesting habitat. Conservation Biology Tignor Miller HL. (eds). Cambridge University Press: 19: 482–491. Cambridge; New York; 748–845. Fish MR, Cote IM, Horrocks JA, Mulligan B, Watkinson AR, Miller J, Limpus CJ. 1981. Incubation period and sexual Jones AP. 2008. Construction setback regulations and differentiation in the green turtle Chelonia mydas. In sea-level rise: mitigating sea turtle nesting beach loss. Proceedings of the Melbourne Herpetological Symposium, Ocean and Coastal Management 51: 330–341. DOI: Banks C, Martin A. (eds). The Zoological Board of Victoria: 10.1016/j.ocecoaman. Melbourne; 66–73. 2007.09.002 Mimura N. 1999. Vulnerability of island countries in the South Fletcher III CH. 1992. Sea-level trends and physical Pacific to sea level rise and climate change. Climate Research consequences: applications to the U.S. shore. Earth-Science 12:137–143. Reviews 33: 73–109. Morreale SJ, Ruiz GJ, Spotila JR, Standora EA. 1982. Girondot M, Tucker AD, Rivalan P, Godfrey MH, Chevalier Temperature-dependent sex determination – current J. 2002. Density-dependent nest destruction and population practices threaten conservation of sea turtles. Science 216: fluctuations of Guianan leatherback turtles. Animal 245–1247. Conservation 5: 75–84. Mwakumanya MA, Bdo O. 2007. Beach morphological Gornitz V. 1991. Global coastal hazards from future sea level dynamics: a case study of Nyali and Bamburi beaches in rise. Global and Planetary Change 3: 379–398. Mombasa, Kenya. Journal of Coastal Research 23: 374–379.

Copyright r 2009 John Wiley & Sons, Ltd. Aquatic Conserv: Mar. Freshw. Ecosyst. (2009) DOI: 10.1002/aqc MMPB FUENTES ET AL.

Nicholls RJ. 2002. Analysis of global impacts of sea-level rise: Snoussi M, Tachfine O, Saida N. 2008. Vulnerability a case study of flooding. Physics and Chemistry of the Earth assessment of the impact of sea-level rise and flooding on 27: 1455–1466. the Moroccan coast: the case of the Mediterranean eastern Nicholls RJ. 2004. Coastal flooding and wetland loss in the zone. Estuarine, Coastal and Shelf Science 77: 206–213. DOI: 21st century: changes under SRES climate and socio- 10.1016/j.ecss.2007.09.024. economic scenarios. Global Environmental Change 14: Titus JG, Park RA, Leatherman SP, Weggel JR. 1991. 69–86. DOI: 10.1016/j.gloenvcha.2003.10.007. Greenhouse effect and sea level rise: the cost of holding Nicholls RJ, Tol RSJ. 2006. Impacts and responses to sea-level back the sea. Coastal Management 19: 171–204. rise: a global analysis of the SRES scenarios over the twenty- Turner M, Batianoff G. 2007. Vulnerability of island flora and first century. Philosophical Transactions of the Royal Society fauna in the Great Barrier Reef to climate change. In A 364: 1073–1095. DOI: 10.1098/rsta.2006.1754 Climate Change and the Great Barrier Reef: A Vulnerability Nicholls RJ, Hoozemans FMJ, Marchand M. 1999. Increasing Assessment, Johnson J, Marshall P (eds). Great Barrier Reef flood risk and wetland losses due to sea-level rise: regional Marine Park Authority and Greenhouse Office: Townsville; and global analysis. Global Environmental Change 9: S69–S87. 621–666. Overpeck JT, Otto-Bliesner BL, Miller GH, Muhs DR, Alley Woodroffe CD. 2008. Reef-island topography and the RB, Kiehl JT. 2006. Paleoclimatic evidence for future ice- vulnerability of atolls to sea-level rise. Global and Planetary sheet instability and rapid sea-level rise. (REPORTS). Change. 62: 77–96. DOI: 10.1016/j.gloplacha.2007.11.001. Science 311: 1747–1750. Woodroffe CD, McLean RF, Smithers SG, Lawson EM. 1999. Ross MS, O’Brien JJ, Sternberg LDSL. 1994. Sea-level rise Atoll reef-island formation and response to sea-level change: and the reduction in pine forests in the Florida keys. West Island, Cocos (Keeling) Islands. Marine Geology 160: Ecological Applications 4: 144–156. 85–104.

Copyright r 2009 John Wiley & Sons, Ltd. Aquatic Conserv: Mar. Freshw. Ecosyst. (2009) DOI: 10.1002/aqc